| Autor: |
Viacheslav Shibaev, Eckehard Olbrich, Jürgen Jost, Ivan P. Yamshchikov |
| Jazyk: |
angličtina |
| Rok vydání: |
2023 |
| Předmět: |
|
| Zdroj: |
Entropy, Vol 25, Iss 2, p 322 (2023) |
| Druh dokumentu: |
article |
| ISSN: |
1099-4300 |
| DOI: |
10.3390/e25020322 |
| Popis: |
A growing number of papers on style transfer for texts rely on information decomposition. The performance of the resulting systems is usually assessed empirically in terms of the output quality or requires laborious experiments. This paper suggests a straightforward information theoretical framework to assess the quality of information decomposition for latent representations in the context of style transfer. Experimenting with several state-of-the-art models, we demonstrate that such estimates could be used as a fast and straightforward health check for the models instead of more laborious empirical experiments. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
|
Nepřihlášeným uživatelům se plný text nezobrazuje |
K zobrazení výsledku je třeba se přihlásit.
|
